A comprehensive shoe quality detection method and system using machine vision

Abstract

The present invention discloses a comprehensive shoe quality detection method and system using machine vision, and relates to the field of image processing technology. The comprehensive shoe quality detection method using machine vision includes: obtaining a water vapor image corresponding to water vapor passing through the upper of a target shoe; identifying a steam column area in the water vapor image according to a pixel parameter value corresponding to each pixel point in the water vapor image; determining the steam column directionality of each top sub-block area in the steam top area using the steam column pixel points in the steam top area of ​​the steam column area, the steam column directionality is used to characterize the degree of deformation of the steam column shape in the top sub-block area; determining the diffusion uniformity of water vapor using the steam column directionality of each top sub-block area, the diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and system for comprehensively detecting shoe body quality using machine vision. Background Art

[0002] In the modern market, the quality of the shoe body directly affects consumers' purchasing decisions and usage experience. As people pay more attention to health and comfort, the quality inspection of the shoe body, especially the breathability inspection, becomes particularly important.

[0003] In the existing method, a measurable water vapor permeability effect is produced on the surface of the shoe body in a specific environment, and an image of the surface of the shoe body is taken. The air permeability of the shoe body is detected by analyzing the area of ​​water vapor permeating the shoe surface in the image.

[0004] However, the existing methods only evaluate the air permeability of the shoe body based on the area of ​​water vapor passing through the shoe upper, and the accuracy of the air permeability detection of the shoe body is low. Summary of the invention

[0005] The embodiments of the present invention provide a method and system for comprehensive shoe quality detection using machine vision, which can improve the accuracy of shoe air permeability detection.

[0006] According to a first aspect of an embodiment of the present invention, a method for comprehensively detecting shoe quality using machine vision is provided, comprising:

[0007] Acquire a water vapor image corresponding to water vapor passing through the upper of a target shoe body;

[0008] Identify the steam column area in the water vapor image according to the pixel parameter value corresponding to each pixel point in the water vapor image;

[0009] Using the steam column pixel points in the steam top area of ​​the steam column area, the steam column orientation of each top block area in the steam top area is determined, and the steam column orientation is used to characterize the degree of deformation of the shape of the steam column in the top block area;

[0010] The vapor column direction of each top block area is used to determine the diffusion uniformity of water vapor, and the diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

[0011] According to a first aspect of an embodiment of the present invention, a system for comprehensively detecting shoe quality using machine vision is provided, comprising:

[0012] An image acquisition module, used to acquire a water vapor image corresponding to water vapor passing through the upper of a target shoe body;

[0013] A region recognition module, used to recognize the steam column region in the water vapor image according to the pixel parameter value corresponding to each pixel point in the water vapor image;

[0014] A direction determination module is used to determine the direction of the steam column of each top block area in the steam top area by using the steam column pixel points in the steam top area of ​​the steam column area, and the steam column direction is used to characterize the degree of deformation of the shape of the steam column in the top block area;

[0015] The uniformity determination module is used to determine the diffusion uniformity of water vapor by using the directionality of the vapor column in each top block area. The diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

[0016] In the comprehensive shoe quality detection method using machine vision provided by the embodiment of the present invention, the vapor column area in the water vapor image is identified based on the pixel parameter value corresponding to each pixel point in the water vapor image. Then, the directionality of the vapor column in the vapor column area is calculated, and the diffusion effect of the water vapor is judged based on the directionality of the vapor column. Then, based on the diffusion effect of the water vapor, the diffusion uniformity of the water vapor is further calculated. Based on the diffusion uniformity of the water vapor, the air permeability of the upper of the target shoe can be evaluated. In this way, the embodiment of the present invention can accurately judge the quality of the shoe air permeability by comprehensively evaluating the vapor column formed after the water vapor passes through the upper, thereby improving the accuracy of the shoe air permeability detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A schematic flow chart of a first method for comprehensive shoe quality inspection using machine vision provided by an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of a water vapor image corresponding to water vapor passing through a shoe upper of a target shoe body provided by an embodiment of the present invention;

[0020] Figure 3 A schematic diagram of a vapor column region provided by one embodiment of the present invention;

[0021] Figure 4 A schematic diagram of edge detection results of a water vapor image provided by an embodiment of the present invention;

[0022] Figure 5 A schematic flow chart of a second method for comprehensively detecting shoe body quality using machine vision provided by an embodiment of the present invention;

[0023] Figure 6 A schematic diagram of a window constructed with a target pixel point as the center provided by an embodiment of the present invention;

[0024] Figure 7 A schematic flow chart of a third method for comprehensively detecting shoe body quality using machine vision provided by an embodiment of the present invention;

[0025] Figure 8 A schematic diagram of a top block area in a vapor column area provided by an embodiment of the present invention;

[0026] Fig. 9 A schematic flow chart of a fourth method for comprehensive shoe quality inspection using machine vision provided by an embodiment of the present invention;

[0027] Fig.10 A schematic diagram of the structure of a comprehensive shoe quality inspection system using machine vision provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a comprehensive shoe quality detection method and system using machine vision proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0029] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0030] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of laws and regulations.

[0031] It should be noted that in the embodiments of the present invention, certain software, components, models and other existing solutions in the industry may be mentioned, which should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present invention, but it does not mean that the applicant has or will necessarily use the solution.

[0032] In the existing method, a measurable water vapor permeability effect is produced on the surface of the shoe body in a specific environment, and the surface image of the shoe body is taken. The air permeability of the shoe body is detected by analyzing the area of ​​water vapor permeating the shoe surface in the image. However, the existing method only evaluates the air permeability of the shoe body based on the area of ​​water vapor permeating the shoe surface, and the accuracy of the air permeability detection of the shoe body is low.

[0033] The purpose of the present invention is to provide a method and system for comprehensive shoe quality detection using machine vision. In the method for comprehensive shoe quality detection using machine vision provided by the embodiment of the present invention, the vapor column area in the water vapor image is identified according to the pixel parameter value corresponding to each pixel point in the water vapor image. Then, the directionality of the vapor column in the vapor column area is calculated, and the diffusion effect of the water vapor is judged according to the directionality of the vapor column. Then, based on the diffusion effect of the water vapor, the degree of diffusion uniformity of the water vapor is further calculated. According to the degree of diffusion uniformity of the water vapor, the air permeability of the upper of the target shoe can be evaluated. In this way, the embodiment of the present invention can accurately judge the quality of the air permeability of the shoe body by comprehensively evaluating the vapor column formed after the water vapor passes through the upper, thereby improving the accuracy of the air permeability detection of the shoe body.

[0034] The following describes a specific embodiment of a method and system for comprehensively detecting shoe body quality using machine vision provided by an embodiment of the present invention.

[0035] Figure 1 A flow chart of a comprehensive shoe quality detection method using machine vision is provided. The comprehensive shoe quality detection method using machine vision can be applied to a server. The comprehensive shoe quality detection method using machine vision can include the following S101 to S104.

[0036] S101, obtaining a water vapor image corresponding to water vapor passing through a shoe upper of a target shoe body.

[0037] In this embodiment, sufficient water is added to a water container and heated to a boiling state to generate water vapor. At the same time, the speed of generating water vapor is adjusted to ensure that it is within a controllable range to avoid excessive water vapor interfering with the experimental results. Then, the target shoe body is placed on the experimental table, ensuring that the upper of the shoe faces upward and the inside of the shoe faces downward to facilitate the entry of water vapor. The water vapor generator is started to allow the water vapor to gradually enter the interior of the target shoe body.

[0038] Then, the server uses a high-resolution camera to capture the water vapor image when the water vapor passes through the upper of the target shoe body under the experimental environment temperature of 20-25℃ and the humidity maintained between 40% and 60%. Figure 2 As shown, a schematic diagram of a water vapor image corresponding to when water vapor passes through the upper of a target shoe body is provided.

[0039] S102, identifying a steam column region in the water vapor image according to a pixel parameter value corresponding to each pixel point in the water vapor image.

[0040] In this embodiment, the pixel parameter value is used to characterize the parameter value of the pixel point for identifying the steam column area in the water vapor image. For example, the pixel parameter value may be the gray value of the pixel point.

[0041] The vapor column area is used to characterize the image area where the vapor column formed by water vapor is located. Figure 3 As shown, a schematic diagram of a vapor column region is provided, wherein 301 is the vapor column region in the water vapor image.

[0042] During the comprehensive test of the air permeability of the target shoe, water vapor, due to its high temperature and large kinetic energy, will diffuse in all directions and gradually fill the interior of the target shoe. Water vapor diffuses outward through the tiny pores or fiber gaps in the upper, and overflows from the interior of the target shoe through the upper. When hot water vapor is released into the air, the temperature of the surrounding air is usually lower. This temperature difference causes the water vapor to cool rapidly and condense to form water droplets. This phenomenon is particularly obvious when the vapor passes through cold air, resulting in enhanced visibility of the vapor column.

[0043] That is, the vapor column area usually appears as a region with a higher grayscale value in the water vapor image, which forms a clear contrast with the surrounding black background and the target shoe surface.

[0044] Therefore, after capturing the water vapor image, the server can further use the edge detection algorithm to perform edge detection on the water vapor image. Specifically, the Sobel edge detection algorithm can be used to perform edge detection on the water vapor image. Figure 4 As shown, a schematic diagram of edge detection results of a water vapor image is provided.

[0045] Then, in the edge detection result of the water vapor image, the vapor column area is separated from the background according to the difference in grayscale values ​​between pixels.

[0046] S103, using the steam column pixel points in the steam top area of ​​the steam column area, determining the steam column directionality of each top block area in the steam top area, where the steam column directionality is used to characterize the degree of deformation of the shape of the steam column in the top block area.

[0047] In this embodiment, the vapor column pixel point is used to characterize the pixel point corresponding to the vapor column in the vapor top area.

[0048] Water vapor diffuses outward through tiny pores or fiber gaps in the upper, and this process is carried out under the combined action of the internal and external pressures of the target shoe. When water vapor diffuses outward from the inside of the target shoe, due to the difference in internal and external pressures, vapor columns are formed. These vapor columns show a certain directionality in the direction perpendicular to the upper of the target shoe.

[0049] In the process of rising, the steam column will gradually disperse, and this phenomenon is closely related to the air permeability of the tiny pores or fiber gaps on the shoe surface. For pores or fiber gaps with poor air permeability, the steam column will disperse faster during the rising process and the height will usually be lower. The higher the degree of deformation of the shape of the steam column, that is, the smaller the direction of the steam column.

[0050] As an example, the server divides the preset steam top area in the steam column area into a plurality of top block areas, each of which contains a certain number of steam column pixel points.

[0051] Then, for each top block area, the direction vector of the vapor column pixel point therein is calculated. Specifically, the direction vector can be obtained by calculating the gradient or difference between the pixel points.

[0052] Then, the distribution of the direction vector is statistically analyzed, and the vapor column directionality is calculated based on the distribution of the direction vector. Specifically, the vapor column directionality can be characterized by calculating the inverse of the standard deviation of the direction vector. The larger the standard deviation, the higher the degree of deformation of the vapor column shape, that is, the smaller the vapor column directionality.

[0053] S104, using the vapor column direction of each top block area, determining the diffusion uniformity of the water vapor, and the diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

[0054] In this embodiment, the server calculates the average value of the vapor column directions of all top block areas as a reference for the overall vapor column direction.

[0055] Then, the deviation of the vapor column direction of each top block area from the reference value is calculated, and the average value of each deviation is calculated to obtain the average value of the deviation. The smaller the average value of the deviation, the closer the shape of the vapor column in each area is to the overall average value, that is, the more uniform the diffusion of water vapor is, and the better the air permeability of the upper of the target shoe body.

[0056] In the comprehensive shoe quality detection method using machine vision provided by the embodiment of the present invention, the vapor column area in the water vapor image is identified based on the pixel parameter value corresponding to each pixel point in the water vapor image. Then, the directionality of the vapor column in the vapor column area is calculated, and the diffusion effect of the water vapor is judged based on the directionality of the vapor column. Then, based on the diffusion effect of the water vapor, the diffusion uniformity of the water vapor is further calculated. Based on the diffusion uniformity of the water vapor, the air permeability of the upper of the target shoe can be evaluated. In this way, the embodiment of the present invention can accurately judge the quality of the shoe air permeability by comprehensively evaluating the vapor column formed after the water vapor passes through the upper, thereby improving the accuracy of the shoe air permeability detection.

[0057] As an optional embodiment, Figure 5 As shown, S102 may specifically include the following S501-S502:

[0058] S501, calculating the vapor column belonging degree of each pixel point according to the pixel parameter value corresponding to each pixel point in the water vapor image, where the vapor column belonging degree is used to characterize the possibility that the pixel point belongs to the vapor column area;

[0059] S502, combining the pixel points whose vapor column attribution degree is greater than a preset attribution degree threshold to form a vapor column region in the water vapor image.

[0060] In this embodiment, the vapor column belonging degree is an indicator to measure whether a pixel point belongs to the vapor column area.

[0061] As an example, the vapor column region generally has a higher grayscale value, so the vapor column attribution degree of the pixel point can be determined according to the grayscale value. Specifically, the grayscale value corresponding to the pixel point can be directly determined as its corresponding vapor column attribution degree.

[0062] Then, the server pre-sets a gray value threshold as a preset attribution threshold, and combines the pixels whose attribution of each steam column is greater than the preset attribution threshold to form a steam column area in the water vapor image.

[0063] Through this embodiment, according to the pixel parameter value corresponding to each pixel in the water vapor image, the vapor column attribution of each pixel can be accurately determined, thereby accurately screening the vapor column area in the water vapor image. In this way, by accurately identifying the vapor column area in the water vapor image, it is helpful to conduct a comprehensive evaluation of the vapor columns in the vapor column area in the subsequent process, and can accurately judge the air permeability of the shoe body, thereby improving the accuracy of the shoe body air permeability detection.

[0064] As an optional embodiment, S501 may specifically include:

[0065] For each pixel in the water vapor image, perform the following steps:

[0066] Taking the target pixel as the center, constructing a target window, a first adjacent window, and a second adjacent window corresponding to the target pixel according to a preset size, wherein the target pixel is any one of the pixels of the water vapor image;

[0067] According to the pixel parameter values ​​corresponding to each pixel point, a target parameter mean in the target window, a first adjacent parameter mean in the first adjacent window, and a second adjacent parameter mean in the second adjacent window are calculated respectively;

[0068] The vapor column attribution degree of the target pixel point is determined by using the target parameter mean, the first adjacent parameter mean and the second adjacent parameter mean.

[0069] In this embodiment, the target window is a window created with the target pixel point as the center according to a preset size. The first adjacent window and the second adjacent window are two outer windows adjacent to the target window.

[0070] The target parameter mean is used to characterize the mean of the pixel parameter values ​​of all pixels in the target window, the first adjacent parameter mean is used to characterize the mean of the pixel parameter values ​​of all pixels in the first adjacent window, and the second adjacent parameter mean is used to characterize the mean of the pixel parameter values ​​of all pixels in the second adjacent window.

[0071] As an example, Figure 6 As shown, a schematic diagram of a window constructed with a target pixel point as the center is provided. The server constructs a target window 602 with a preset size of 3x3 with a target pixel point 601 as the center. At the same time, a first adjacent window 603 is constructed on the left side of the target window 602 with a preset size of 3x3, and a second adjacent window 604 is constructed on the right side of the target window 602 with a preset size of 3x3.

[0072] Then, for the target window 602 , the first adjacent window 603 , and the second adjacent window 604 , the corresponding target parameter mean, the first adjacent parameter mean, and the second adjacent parameter mean are calculated according to the parameter values ​​of the pixels in the windows.

[0073] Then, the inverse of the difference between the mean of the target parameter and the mean of the first adjacent parameter is set as the first weight, and the inverse of the difference between the mean of the target parameter and the mean of the second adjacent parameter is set as the second weight. According to the first weight and the second weight, the first adjacent parameter mean and the second adjacent parameter mean are weighted and summed to obtain the vapor column attribution degree of the target pixel point.

[0074] Through this embodiment, a target window, a first adjacent window, and a second adjacent window corresponding to a target pixel are created to calculate the vapor column attribution degree of the target pixel, thereby facilitating accurate identification of the vapor column region in the water vapor image, facilitating subsequent comprehensive evaluation of the vapor columns in the vapor column region, and accurately determining the air permeability of the shoe body, thereby improving the accuracy of the shoe body air permeability detection.

[0075] As an optional embodiment, the target parameter mean includes a target gradient mean and a target grayscale mean, the first adjacent parameter mean includes a first adjacent gradient mean and a first adjacent grayscale mean, and the second adjacent parameter mean includes a second adjacent gradient mean and a second adjacent grayscale mean;

[0076] Determining the vapor column attribution degree of the target pixel point by using the target parameter mean, the first adjacent parameter mean, and the second adjacent parameter mean may specifically include:

[0077] Determine the relative intensity of the gradient of the target pixel using the target gradient mean, the first adjacent gradient mean, and the second adjacent gradient mean;

[0078] Determine the relative gray intensity of the target pixel using the target gray mean, the first adjacent gray mean, and the second adjacent gray mean;

[0079] The gradient relative intensity is multiplied by the grayscale relative intensity to obtain the vapor column attribution degree of the target pixel.

[0080] In this embodiment, the vapor column attribution degree of the target pixel point can be specifically determined by the following formula 1:

[0081] Formula 1;

[0082] In the formula, Indicates The vapor column attribution degree of each target pixel. Indicates The target gradient mean of target pixels, Indicates The first neighboring gradient mean of the target pixel, Indicates The second adjacent gradient mean of the target pixel. Indicates The target grayscale mean of target pixels, Indicates The first neighboring grayscale mean of the target pixel, Indicates The second adjacent grayscale mean of the target pixels.

[0083] in, Indicates The relative gradient strength of the target pixel. The larger the relative gradient strength, the greater the The greater the probability that the target pixel belongs to the vapor column area in the water vapor image; Indicates The relative gray intensity of the target pixel point. The larger the relative gray intensity, the greater the The greater the probability that the target pixel belongs to the vapor column area in the water vapor image.

[0084] No. The vapor column attribution degree of the target pixel The larger the value, the The greater the probability that a target pixel belongs to the vapor column area in the water vapor image, the greater the probability that the vapor column belongs to the water vapor image. The vapor column attribution degree is determined by the gradient value and grayscale value of the pixel.

[0085] Through this embodiment, the vapor column attribution of the target pixel can be accurately determined by using the target parameter mean, the first adjacent parameter mean, and the second adjacent parameter mean. In this way, according to the vapor column attribution of each pixel, it is helpful to accurately identify the vapor column area in the water vapor image, so as to accurately judge the air permeability of the shoe body and improve the accuracy of the shoe body air permeability detection.

[0086] As an optional embodiment, Figure 7 As shown, S103 may specifically include the following S701-S703:

[0087] S701, obtaining a vapor top region in a vapor column region;

[0088] S702, dividing the steam top region according to a preset division strategy to obtain a plurality of top block regions in the steam top region;

[0089] S703, for each top block area, determine the steam column directionality of the top block area by using the number of steam column pixels in the top block area, the distance between the steam column pixels, the steam column attribution of the steam column pixels, and the area of ​​the top block area.

[0090] In this embodiment, if Figure 8 As shown, a schematic diagram of the top block area in the steam column area is provided. Among them, after the server uses the Sobel edge detection algorithm to perform edge detection on the water vapor image and identifies the steam column area 801. In the steam column area 801, the steam top area 802 is identified according to the morphological characteristics (such as height, width, etc.) of the steam column.

[0091] Then, the server formulates a suitable preset division strategy according to the characteristics and processing requirements of the steam column. The preset division strategy may be related to parameters such as region size, region shape, and region number. And according to the preset division strategy, the steam top region 802 is divided into a plurality of top block regions 803.

[0092] Then, for each top block area, the number of vapor column pixels in the top block area is counted, which helps to understand the density of the vapor column in the top block area. The distance between the vapor column pixels in the top block area is calculated, which can reflect the distribution uniformity and directionality of the vapor column in the area. The area of ​​the top block area is obtained for normalization or weight distribution.

[0093] Finally, based on the number of vapor column pixels in the top block area, the distance between the vapor column pixels, the vapor column attribution of the vapor column pixels, and the area of ​​the top block area, a preset vapor column directionality algorithm (such as weighted average or cluster analysis) is used to calculate the vapor column directionality of each top block area.

[0094] Through this embodiment, the vapor column direction of each top block area in the vapor column area is calculated according to the vapor column pixel points in the vapor top area. In this way, by accurately calculating the vapor column direction of each top block area in the vapor top area, the diffusion effect of water vapor can be accurately determined. In addition, the air permeability of the shoe body can be accurately determined, thereby improving the accuracy of the shoe body air permeability detection.

[0095] As an optional embodiment, S703 may specifically include:

[0096] The number of vapor column pixels in the top block region is divided by the area of ​​the top block region to obtain the pixel density of the top block region;

[0097] The target interval distance of each steam column pixel point in the top block area is multiplied by the corresponding steam column attribution degree and then accumulated to obtain a target calculation value. The target interval distance is used to represent the average value of the distance between the steam column pixel point and the surrounding adjacent steam column pixel points;

[0098] The pixel density of the top block area is multiplied by the target calculation value to obtain the vapor column directionality of the top block area.

[0099] In this embodiment, the directionality of the steam column in the top block area can be specifically determined by the following formula 2:

[0100] Formula 2;

[0101] In the formula, Indicates the direction of the vapor column in the a-th top block area, Indicates the number of vapor column pixels in the a-th top block area, Represents the area of ​​the a-th top block. Indicates the ath top block area The target spacing distance of the steam column pixels, specifically, the The average of the distances between the steam column pixel points and the nearest steam column pixel points in each direction is calculated. Indicates the ath top block area The vapor column attribution degree of each vapor column pixel.

[0102] in, Indicates the pixel density of the a-th top block area. The more the number of vapor column pixels is, the smaller the area of ​​the top block is, the greater the pixel density is, the top of the vapor column has not yet dispersed, and the corresponding vapor column direction is greater. Considering the distance of each vapor column pixel and the vapor column attribution, a weighted distance value can be obtained by multiplying the two. The larger this value is, the greater the distance between a certain vapor column pixel and its adjacent pixels is, which may indicate that the pixel occupies an important position in the overall vapor column shape and contributes more to the vapor column directionality.

[0103] The direction of the vapor column in the a-th top block area The larger it is, the more concentrated the top of the vapor column is, the less the shape of the vapor column is deformed, and the better the air permeability is.

[0104] Through this embodiment, according to the number of steam column pixels in the top block area, the distance between the steam column pixels, the steam column attribution of the steam column pixels, and the area of ​​the top block area, the steam column direction of the top block area can be accurately determined. In this way, by accurately calculating the steam column direction of each top block area in the steam top area, the diffusion effect of water vapor can be accurately determined. In addition, the air permeability of the shoe body can be accurately determined, thereby improving the accuracy of the shoe body air permeability detection.

[0105] As an optional embodiment, S104 may specifically include:

[0106] The average value of the steam column directionality of each top block area is calculated to obtain the average value of the steam column directionality;

[0107] The difference between the vapor column directionality of each top block area and the average of the vapor column directionality is averaged to obtain the vapor column consistency;

[0108] Using the vapor column pixel points in each top block area, determining the standard deviation of the vapor column height between each top block area;

[0109] Multiplying the vapor column consistency by the standard deviation of the vapor column height gives the uniformity of water vapor diffusion.

[0110] In this embodiment, the diffusion uniformity of water vapor can be specifically determined by the following formula 3:

[0111] Formula 3;

[0112] In the formula, Indicates the uniformity of water vapor diffusion. Indicates the direction of the vapor column in the a-th top block area, represents the mean value of the vapor column direction of each top block area, A represents the number of top block areas, and H represents the standard deviation of the vapor column height between each top block area.

[0113] in, Indicates the consistency of the direction of the vapor column in the vapor column area. The larger the value, the greater the consistency of the direction, and the more uniform the diffusion of water vapor. The larger the standard deviation H of the vapor column height, the more uniform the height distribution of the vapor column, and the more uniform the diffusion of water vapor.

[0114] The greater the diffusion uniformity K of water vapor, the more uniform the ventilation of the target shoe body.

[0115] Through this embodiment, the diffusion uniformity of water vapor is determined according to the direction of the vapor column of each top block area. In this way, the air permeability of the upper of the target shoe body is evaluated according to the diffusion uniformity of the water vapor, and the air permeability of the shoe body can be accurately judged, thereby improving the accuracy of the shoe body air permeability detection.

[0116] As an optional embodiment, the water vapor image includes a plurality of images;

[0117] like Fig. 9 As shown, after S104, the comprehensive shoe body quality inspection method using machine vision may further include the following S901:

[0118] S901, calculating the air permeability evaluation degree of the target shoe body according to the diffusion uniformity corresponding to each water vapor image.

[0119] In this embodiment, after obtaining the diffusion uniformity corresponding to each water vapor image, the server determines the maximum and minimum values ​​of the diffusion uniformity in all images. Then, the diffusion uniformity of each image is normalized to a range of 0 to 1. Specifically, the diffusion uniformity of the target image is normalized by dividing the difference between the diffusion uniformity of the target image and the minimum value of the diffusion uniformity by the difference between the maximum value of the diffusion uniformity and the minimum value of the diffusion uniformity.

[0120] Finally, the average value of the normalized diffusion uniformity is calculated to obtain the air permeability evaluation degree of the target shoe body.

[0121] Through this embodiment, the diffusion uniformity corresponding to multiple water vapor images is used to calculate the air permeability evaluation degree of the target shoe body, which can avoid the existence of errors, thereby accurately judging the air permeability of the shoe body and improving the accuracy of the shoe body air permeability detection.

[0122] As an optional embodiment, S901 may specifically include:

[0123] Obtaining a maximum diffusion uniformity value and a minimum diffusion uniformity value from each diffusion uniformity;

[0124] For each diffusion uniformity, the difference between the diffusion uniformity and the minimum diffusion uniformity is divided by the difference between the maximum diffusion uniformity and the minimum diffusion uniformity to obtain the normalized uniformity corresponding to each diffusion uniformity;

[0125] The average of each normalized uniformity is calculated to obtain the air permeability evaluation degree of the target shoe body.

[0126] In this embodiment, the air permeability evaluation degree of the target shoe body can be specifically determined by the following formula 4:

[0127] Formula 4;

[0128] In the formula, T represents the air permeability evaluation degree of the target shoe body, M represents the number of water vapor images, Indicates Diffusion uniformity of a water vapor image. Indicates that when the water vapor image is In the interval The minimum value of Indicates that when the water vapor image is In the interval The maximum value of .

[0129] in, Indicates the normalization of the diffusion uniformity of each water vapor image. The average value of the normalized diffusion uniformity of each water vapor image is calculated.

[0130] For example, after obtaining the air permeability evaluation degree of the target shoe body, a permeability evaluation degree threshold value of 0.5 can be preset. When the air permeability evaluation degree of the target shoe body is greater than or equal to 0.5, it means that the air permeability of the target shoe body is good and the quality inspection of the target shoe body is qualified; when the air permeability evaluation degree of the target shoe body is less than 0.5, it means that the air permeability of the target shoe body is poor and the quality inspection of the target shoe body is unqualified.

[0131] According to this embodiment, the diffusion uniformity corresponding to multiple water vapor images is normalized and then the average is calculated to obtain the air permeability evaluation degree of the target shoe body, which can avoid the existence of errors, thereby accurately judging the air permeability of the shoe body and improving the accuracy of the shoe body air permeability detection.

[0132] A comprehensive shoe quality detection method based on machine vision. Accordingly, the present invention also provides a specific embodiment of a comprehensive shoe quality detection system using machine vision.

[0133] Fig.10 A structural schematic diagram of a comprehensive shoe body quality inspection system using machine vision provided in an embodiment of the present application is shown. The comprehensive shoe body quality inspection system using machine vision 1000 may include an image acquisition module 1010, a region recognition module 1020, a direction determination module 1030 and a uniformity determination module 1040.

[0134] An image acquisition module 1010 is used to acquire a water vapor image corresponding to water vapor passing through the upper of a target shoe body;

[0135] The region identification module 1020 is used to identify the steam column region in the water vapor image according to the pixel parameter value corresponding to each pixel point in the water vapor image;

[0136] A directionality determination module 1030 is used to determine the directionality of the steam column of each top block area in the steam top area by using the steam column pixel points in the steam top area of ​​the steam column area, and the steam column directionality is used to characterize the degree of deformation of the shape of the steam column in the top block area;

[0137] The uniformity determination module 1040 is used to determine the diffusion uniformity of water vapor using the directionality of the vapor column of each top block area. The diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

[0138] In the comprehensive shoe quality detection system using machine vision provided by the embodiment of the present invention, the vapor column area in the water vapor image is identified according to the pixel parameter value corresponding to each pixel point in the water vapor image. Then, the directionality of the vapor column in the vapor column area is calculated, and the diffusion effect of the water vapor is judged according to the directionality of the vapor column. Then, based on the diffusion effect of the water vapor, the diffusion uniformity of the water vapor is further calculated. According to the diffusion uniformity of the water vapor, the air permeability of the upper of the target shoe can be evaluated. In this way, the embodiment of the present invention can accurately judge the quality of the shoe air permeability by comprehensively evaluating the vapor column formed after the water vapor passes through the upper, thereby improving the accuracy of the shoe air permeability detection.

[0139] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0140] It should also be noted that the exemplary embodiments mentioned in the present invention describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in a different order from the embodiments, or several steps can be performed simultaneously.

[0141] The above is only a specific implementation of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be covered within the protection scope of the present invention.

Claims

1. A method for comprehensive shoe quality inspection using machine vision, characterized in that: The method comprises: Acquire a water vapor image corresponding to water vapor passing through the upper of a target shoe body; identifying a vapor column region in the water vapor image according to a pixel parameter value corresponding to each pixel point in the water vapor image; Determine the steam column directionality of each top sub-block area in the steam top area by using the steam column pixel points in the steam top area of ​​the steam column area, wherein the steam column directionality is used to characterize the degree of deformation of the shape of the steam column in the top sub-block area; The diffusion uniformity of the water vapor is determined by using the directionality of the vapor column of each of the top block areas, and the diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

2. The method for comprehensive shoe quality inspection using machine vision according to claim 1 is characterized in that: The step of identifying a vapor column region in the water vapor image according to a pixel parameter value corresponding to each pixel point in the water vapor image comprises: Calculating the vapor column belonging degree of each pixel point according to the pixel parameter value corresponding to each pixel point in the water vapor image, wherein the vapor column belonging degree is used to characterize the possibility that the pixel point belongs to the vapor column area; The pixel points whose vapor column attribution degree is greater than a preset attribution degree threshold are combined to form a vapor column region in the water vapor image.

3. The method for comprehensive shoe quality inspection using machine vision according to claim 2, characterized in that: The calculating the vapor column attribution degree of each pixel point according to the pixel parameter value corresponding to each pixel point in the water vapor image comprises: For each pixel in the water vapor image, the following steps are performed respectively: Taking a target pixel point as the center, constructing a target window, a first adjacent window, and a second adjacent window corresponding to the target pixel point according to a preset size, wherein the target pixel point is any one of the pixel points of the water vapor image; According to the pixel parameter values ​​corresponding to the pixel points, a target parameter mean in the target window, a first adjacent parameter mean in the first adjacent window, and a second adjacent parameter mean in the second adjacent window are calculated respectively; The vapor column attribution degree of the target pixel point is determined by using the target parameter mean, the first adjacent parameter mean, and the second adjacent parameter mean.

4. The method for comprehensive shoe quality inspection using machine vision according to claim 3, characterized in that: The target parameter mean includes a target gradient mean and a target grayscale mean, the first adjacent parameter means include a first adjacent gradient mean and a first adjacent grayscale mean, and the second adjacent parameter means include a second adjacent gradient mean and a second adjacent grayscale mean; The step of determining the vapor column attribution degree of the target pixel point by using the target parameter mean, the first adjacent parameter mean, and the second adjacent parameter mean includes: Determine the relative gradient strength of the target pixel by using the target gradient mean, the first adjacent gradient mean, and the second adjacent gradient mean; Determine the relative gray intensity of the target pixel by using the target gray mean, the first adjacent gray mean and the second adjacent gray mean; The gradient relative intensity is multiplied by the grayscale relative intensity to obtain the vapor column attribution degree of the target pixel.

5. The method for comprehensive shoe quality inspection using machine vision according to claim 1, characterized in that: The method of using the steam column pixel points in the steam top region of the steam column region to determine the steam column directionality of each top block region in the steam top region includes: obtaining a vapor top region in the vapor column region; Dividing the steam top region according to a preset division strategy to obtain a plurality of top block regions in the steam top region; For each of the top block areas, the vapor column directionality of the top block area is determined by using the number of vapor column pixel points in the top block area, the distance between the vapor column pixel points, the vapor column attribution degree of the vapor column pixel points, and the area of ​​the top block area.

6. The method for comprehensive shoe quality inspection using machine vision according to claim 5, characterized in that: For each of the top block regions, the steam column directionality of the top block region is determined by using the number of steam column pixels in the top block region, the distance between the steam column pixels, the steam column attribution of the steam column pixels, and the area of ​​the top block region, including: Dividing the number of vapor column pixels in the top block region by the area of ​​the top block region to obtain the pixel density of the top block region; The target interval distance of each steam column pixel point in the top block area is multiplied by the corresponding steam column attribution degree and then accumulated to obtain a target calculation value, wherein the target interval distance is used to represent the average value of the distance between the steam column pixel point and the surrounding adjacent steam column pixel points; The pixel density of the top block area is multiplied by the target calculation value to obtain the vapor column directionality of the top block area.

7. The method for comprehensive shoe quality inspection using machine vision according to claim 1, characterized in that: Determining the diffusion uniformity of the water vapor by using the vapor column direction of each of the top block areas includes: Calculating the average value of the steam column directionality of each of the top block areas to obtain the average value of the steam column directionality; Calculating the average of the differences between the vapor column directions of each of the top block areas and the average of the vapor column directions to obtain vapor column consistency; Determine the standard deviation of the height of the steam column between the top sub-block areas by using the steam column pixel points in the top sub-block areas; The uniformity of diffusion of the water vapor is obtained by multiplying the vapor column consistency by the standard deviation of the vapor column height.

8. The method for comprehensive shoe quality inspection using machine vision according to any one of claims 1 to 7, characterized in that: The water vapor images include a plurality of images; After determining the diffusion uniformity of the water vapor by using the vapor column direction of each of the top block areas, the method further includes: The air permeability evaluation degree of the target shoe body is calculated according to the diffusion uniformity corresponding to each of the water vapor images.

9. The method for comprehensive shoe quality inspection using machine vision according to claim 8, characterized in that: The calculating the air permeability evaluation degree of the target shoe body according to the diffusion uniformity corresponding to each of the water vapor images respectively includes: Obtaining a maximum diffusion uniformity value and a minimum diffusion uniformity value from each of the diffusion uniformities; For each of the diffusion uniformities, the difference between the diffusion uniformity and the minimum diffusion uniformity is divided by the difference between the maximum diffusion uniformity and the minimum diffusion uniformity to obtain a normalized uniformity corresponding to each of the diffusion uniformities; The average of each normalized uniformity is calculated to obtain the air permeability evaluation degree of the target shoe body.

10. A comprehensive shoe quality inspection system using machine vision, characterized in that: The system comprises: An image acquisition module, used to acquire a water vapor image corresponding to water vapor passing through the upper of a target shoe body; A region identification module, used to identify the steam column region in the water vapor image according to the pixel parameter value corresponding to each pixel point in the water vapor image; A directionality determination module, used to determine the directionality of the steam column of each top block area in the steam top area by using the steam column pixel points in the steam top area of ​​the steam column area, wherein the steam column directionality is used to characterize the degree of deformation of the shape of the steam column in the top block area; The uniformity determination module is used to determine the diffusion uniformity of the water vapor by using the directionality of the vapor column of each of the top block areas, and the diffusion uniformity is used to evaluate the air permeability of the upper of the target shoe body.

Images